• Keine Ergebnisse gefunden

Solubility equilibria and metastable zone width determination

4. Results and discussion

4.2 Solubility equilibria and metastable zone width determination

Ch. 4 Results and discussion

(eq. 4.6) in the crystallizer as well as total and recovered crystal product mass (eqs. 4.7 and 4.8 respectively) with time were derived. In order to use the proposed model there are parameters, which should be determined experimentally. Concerning the physical properties (e.g. crystal density) of the substance, these are already known and can be taken from the literature. On the contrary, some parameters like substance solubility are more complex and depend on the physical and chemical properties of the solute and solvent as well as on temperature, pressure and the pH of the solution.

To quantify the studied process in particular, profound knowledge of the solubility equilibria along with the metastable zone width (MSZW) of the substance to be crystallized is needed.

4.2 Solubility equilibria and metastable zone width determination

Results and discussion Ch. 4

It can be seen from the figure, a positive slope of the solubility curves is observed, i.e. the solubility of the species increases in an exponential dependence on the temperature. The measurements of L-glu and D-glu solutions fit almost perfectly to literature data and show identical progression. On the contrary, the literature data presented for DL-glu show a considerable difference in absolute values. In a comparison of the solubility data of the pure enantiomers and DL-glu.H2O data from Dalton and Schmidt, it can be noted that the solubility of the DL-form is greater than the corresponding values for the enantiomer form.

They also reported that they have obtained DL-glu.H2O by racemizing D-glu with barium hydroxide in an autoclave [Dalton1933]. Our solubility measurements of DL-glu.H2O are in very good agreement with the results from Dalton and Schmidt, while the results published by Apelblat for the glutamic acid racemate are used later for constructing the Van’t Hoff plot. A single probe of an aqueous solution of both enantiomers with 50:50 ratio was measured at 50°C by using an isothermal method (described in chapter 3.3.1.1). The result for the measured by HPLC saturation concentration (3.91 wt.%) fit almost perfectly with the data from Apelblat.

As seen in fig. 4.2, the solubilities of the enantiomers are lower than the solubilities of the racemate in the glutamic acid system. For the ideal case, the solubility of the conglomerate racemic solution equals the sum of the solubilities of the single enantiomers. In our case, with increase of temperature, the solubility ratios between the racemic mixture (data from Apelblat) and the pure enantiomers change from ~1.78 at 10°C to ~1.84 at 65°C. A comparison with solubility data from Dalton and our measurements of DL-glu.H2O with the glutamic acid enantiomers show ratios from 2.45 at 10°C to 2.19 at 65°C between the DL-form and the pure enantiomers. The van’t Hoff plot (shown in Fig. 4.3) gives a good linearization and the slopes for the 3 species are very similar.

The trend lines, shown in figure 4.3, correspond to the solubility data of DL-glu, D-glu and L-glu, and are represented as a natural logarithm of the molar concentration of the respective substance versus the absolute reciprocal temperature. Based on the data, the van’t Hoff equation can be written by using

Ch. 4 Results and discussion

Fig. 4.3 Van’t Hoff plot of D-glu, L-glu and DL-glu.

ln(xeq) = A.T-1 + B

Table 4.1 Parameters for the equation 4.16 for the glutamic acid system.

Substance A = ΔH / R B = ΔS / R DL-glu.H2O

D-glu L-glu

- 3546.6 - 3764.6 - 3945.8

5.8622 5.7296 6.2774

In the table, ΔH / ΔS represent the change of the molar enthalpy / entropy in the solution over a temperature range and R is the gas constant. Based on the parameter A estimated (the slope of the respected trend line), the molar enthalpies of solution can be calculated from the equation 4.18. The estimated molar enthalpies for DL-glu, L-glu and D-glu are shown in table 4.2 and compared with the literature data.

Table 4.2 Calculated molar enthalpies ΔH (kJ/mol) for the glutamic acid system, compared with literature data measurements at 298.15 K.

Substance this work Apelblat1997 Dalton1933 DL-glu

D-glu L-glu

29.5*

31.3 32.8

27.3 - 30.2

23.9 – 25.9*

25.4 25.3

* The molar enthalpies shown are for DL-glu.H2O

From the table it can be seen that the calculated values are in fair agreement.

Still, our results are more close to these, reported by Apelblat et al.

(4.16)

Results and discussion Ch. 4

Essential information, needed for the optimal realization of the crystallization, is the estimation of the metastable zone width. In Fig. 4.4 are depicted the results from the polythermal experiments, done with Crystal 16TM.

The MSZW of the single enantiomers is broad with a ΔTmax > 20 K between the solubility and supersolubility curves for saturation concentrations below 55°C.

From the measurements it can be concluded that crystallization experiments can be conducted in relatively high supersaturated solutions with a considerable increase in productivity. Hence, for the crystallization experiments of L-glu, saturation concentration at 50°C and a moderate supersaturation of 1.19 (Tc = 45°C) were used and the results are shown in chapter 4.4.

Fig. 4.4 MSZW of L-glu. The solubility / supersolubility curves are depicted with solid / dashed lines respectively correspondent to the appropriate polynomial equation. Experimental data for solubility / supersolubility is given with circles / triangles. The data for D-glu overlaps with L-glu data in the entire temperature region shown and it is not depicted.

The MSZW of DL-glu.H2O was also measured, but the results are not represented. As DL-glu.H2O is a racemic compound, it cannot be separated into single enantiomers by preferential crystallization [Yokota2006]. It can also be

Ch. 4 Results and discussion

difference between the measured samples of DL-glu.H2O and the pure enantiomers.

- Solubility equilibria and MSZW of the asparagine monohydrate system

The solubilities of D-asn.H2O, L-asn.H2O and DL-asn.H2O in water were determined gravimetrically in the temperature range 10°C – 60°C. In the figure 4.5 are shown the data from the solubility measurements along with literature values, plotted for comparison.

Fig. 4.5 Solubility of the enantiomers and racemic mixtures of asparagine monohydrate as a function of temperature (wt.% - weight percent) (own and literature data).

The solubility courses for both enantiomers D- and L –asn.H2O are almost identical and the dissolved mass increase exponentially with increasing temperature. From the data it can be concluded that the solubility of both enantiomers in water is equal as expected. The literature values are measured in a range from 15 °C to 45 °C and are in very good agreement with our measurements. Notable deviations from the exponential course can be seen at temperatures above 50 °C. This may be due to the evaporation of the water at high temperatures, thus increasing slightly the measured values. Similar with the glutamic acid system, the solubility of the DL-asn.H2O exceeds that for the pure enantiomers. The measured values show almost double magnitude as these for

Results and discussion Ch. 4

the single enantiomers. The construction of the van’t Hoff plot (shown in fig. 4.6) gives almost perfect linearization.

Fig. 4.6 Van’t Hoff plot of D-asn.H2O, L-asn.H2O and DL-asn.H2O.

Based on the isothermal solubility measurements of samples, containing defined amounts of D-Asn.H2O and L-Asn.H2O in water, the ternary solubility phase diagram of the system D-Asn.H2O / L-Asn.H2O / water is constructed with solubility isotherms at 20 and 40 °C, (blue and red lines respectively, see Fig.

4.7).

It can be seen form Fig. 4.7, the ternary system D-Asn.H2O / L-Asn.H2O / water exhibits symmetrical behavior with highest solubility at the eutectic. This means that the aqueous solutions of the racemate possess the highest solubility at the eutectic. The initial concentration of the samples for the 20 °C isotherm determination is given in Fig. 4.7 with green dots. It can be seen, that after establishment of equilibria, three of resulted solutions (blue dots) are racemic or very close to racemic, while in the other three the ratio of the L-enantiomer gradually increases while the respective solubility decreases. The same can be observed for the measured at 40 °C samples, where four from the total six starting solutions (black points) resulted after equilibria in the eutectic. However, by the determination of the 50 °C isotherm, the initial concentrations are closer to the measured ones, thus only three samples are measured at the eutectic.

Ch. 4 Results and discussion

Fig. 4.7 Up: Ternary solubility phase diagram of the system D-Asn.H2O / L-Asn.H2O / water; Down: Upper 20% section of the diagram, shown for clarity.

Isotherms at 20 and 40 °C are shown with blue and red lines respectively.

Additionally, the respective colored dashed lines represent tie-lines, separating two-phase from three-phase zone. Green and black dots represent initial concentration at 20 and 40 °C respectively. Literature data is represented with pink dots [Seebach2011]. Axes are in weight fractions.

0.4

0.96

L-Asn.H2O

0.16 L-Asn.H2O

0.92

water

0.88 D-Asn.H2O

L-Asn.H2O 0.2

water

0.6

D-Asn.H2O

0.8

0.04 0.8

0.08 0.2

0.12 water

0.84

0.4 0.6

Results and discussion Ch. 4

This behavior in the solubility phase diagram concludes that the ternary system D-Asn.H2O / L-Asn.H2O / water exhibit simple eutectic. Moreover, XRPD diffractogram of the samples from the solid phase with racemic ratio, taken at 20 and 40 °C unambiguously show that a mechanical mixture of both substances has been formed. The XRPD diffractogram is given in Fig. A2.2.

In order to optimally perform a preferential crystallization experiment, the metastable zone width of the asparagine racemic solution was studied. The results from the polythermal measurements for determination of the MSZW of DL-asn.H2O are shown in figure 4.8.

Fig. 4.8 Solubility and MSZW of DL-asn.H2O. The solubility / supersolubility curves are depicted with solid / dashed lines respectively correspondent to the appropriate polynomial equation. Experimental data for solubility / supersolubility is given with circles / triangles.

The two curves define an expanding metastable region, where ΔT at concentration of 5 wt.% is about 7 K and rises up to 14 K at 23 wt.%. This suggests that subsequent to the crystallization experiments, a relatively high supersaturation levels could be used. Nevertheless, in high supersaturated solutions the possibility of spontaneous nucleation cannot be neglected. For this purpose, a temperature difference of no more than 8 K appears to be suitable for eventual crystallization experiments.

Ch. 4 Results and discussion

- Solubility equilibria and MSZW of the aminobenzoic acid system

The measurements of the solubility and the metastable zone width (MSZW) of the aminobenzoic acid system were performed for two out of the three possible stereoisomers – ortho-aminobenzoic acid (OABA) and para-aminobenzoic acid (PABA). The polythermal solubility measurements of OABA are done in temperature range 10 – 50 °C (blue dashes in Fig. 4.9). Isothermal solubility measurements of OABA and PABA (in Fig. 4.9 green points and blue crosses respectively) are done at temperatures 20, 35, 40, 45 and 50 °C. The results are shown along with literature data. Solubility of meta-aminobenzoic acid (MABA) is also shown for comparison. It can be clearly seen from the figure, that all three isomers exhibit very low solubility in water, which at 50 °C is between 1.2 and 1.4 wt.%. The pure isomers, having a hydrophobic benzene core, cannot form strong interactions with the water, which explains their low solubility in it. On the contrary, the carboxyl- and amino- groups can form relatively strong hydrogen bonding interactions, thus promoting the solvation of the molecules in the water.

This is confirmed by He et al., which has published theoretical and experimental studies of water complexes of OABA and PABA [He2005].

Fig. 4.9 Solubility of OABA, MABA and PABA aqueous solutions as a function of temperature (wt.% - weight percent) (own and literature data).

Results and discussion Ch. 4

Hence, the measured solubility values show noticeable differences in temperature range below 15 °C and above 35 °C. It must be noted that all three isomers of aminobenzoic acid exhibit polymorphic transformations. The data for PABA shows that both polymorphs are enantiotropic with a transition temperature of 25 °C. Above this temperature the solubility of the alpha-form becomes lower than the beta-form and it crystallizes preferentially [Gracin2004]. The MABA has two polymorphs, form I is the stable and form II – metastable. The data represented in the figure is for the stable form I. Svärd et al. have investigated both MABA polymorphs and have shown that metastable form II turns very quick to form I in aqueous and methanol solutions [Svärd2010]. Almost similar tendency is published for the OABA polymorphs. Jiang et al. have observed that at lower temperature the solubility differences between all forms in water are relative large, while above 50°C the solubility of form I approaches those of the other two forms. The authors have proven that the transformation temperatures of form III to form I are probably lower than that of form II to form I. The transformation results unambiguously show that below 50 °C forms II and III are transforming to form I [Jiang2010]. It can be concluded that form I is the most stable form, while forms II and III are metastable below 50 °C. In our case, only form I of OABA was crystallized and observed microscopically. For PABA, α-form was observed with no appearance of β-form at higher temperatures (above 35°C) and at temperatures below 25°C only α-form was crystallized and observed. The results from XRPD measurements from the respected crystal samples have unambiguously shown, that a polymorph type I of OABA and α-polymorph of PABA have been formed.

Based on the isothermal solubility measurements of samples, containing defined amounts of OABA and PABA in water, the ternary solubility phase diagram of the system OABA / PABA / water is constructed with solubility isotherms at 25, 35 and 50 °C, (blue, green and red lines respectively, see Fig. 4.10).

It can be clearly seen form Fig. 4.10 the ternary system OABA / PABA / water exhibits also symmetrical behavior like the case of asparagine enantiomers in water. In this case, aqueous solutions of OABA and PABA with 50:50 ratios possess the highest solubility at the eutectic.

Ch. 4 Results and discussion

Fig. 4.10 Up: Ternary solubility phase diagram of the system OABA / PABA / water; Down: Upper 10% section of the diagram, shown for clarity. Isotherms at 25, 35 and 50 °C are shown with thick blue, green and red lines respectively.

Additionally, the respective colored thin lines represent tie-lines, separating two-phase from three-two-phase zone. Black dots represent initial concentration. Axes are in weight fractions.

0.5

0.98

PABA

0.08 PABA

0.96

water

0.94 OABA

PABA 0.25

water

0.5

OABA

0.75

0.02 0.75

0.04 0.25

0.06 water

0.92

Results and discussion Ch. 4

The initial concentration of the samples for each isotherm determination was the same and it can be seen, that for the determination of the 20 °C isotherm, all concentration measurements at equilibria are at the eutectic point. Hence, the respective tie-lines delimitate all initial conditions into the three-phase zone of the ternary phase diagram. The same can be observed for the measured at 35 °C samples, where only one point is observed in the respective two-phase region.

However, by the determination of the 50 °C isotherm, the initial concentrations are closer to the measured ones, thus only three samples are measured at the eutectic. This behavior in the solubility phase diagram concludes that the ternary system OABA / PABA / water exhibit simple eutectic. Moreover, XRPD diffractogram of the samples from the solid phase with 1:1 ratio, taken at 20 and 50 °C unambiguously show that a mechanical mixture of both substances has been formed and no polymorphic transformation is observed.

The application of the preferential crystallization to separate one isomer in the presence of the other in this case should be possible. Therefore, the metastable zone width of a solution of both substances was studied. The results from the solubility measurements of a 50:50 aqueous solution of OABA and PABA are shown in Fig. 4.11.

Fig. 4.11 Solubility and MSZW of a 50:50 aqueous solution of OABA and PABA.

The solubility / supersolubility curves are depicted with solid / dashed lines

Ch. 4 Results and discussion

A comparison between the measured data for the pure substances (Fig. 4.9) and the values shown in Fig. 4.11, it can be clearly seen, that the solubilities of the aqueous solutions of the both substances are twice as high as these for the single substance. The estimated broad metastable zone width has a ΔT of about 18 K at lower temperatures and a value of 15 K at 50 °C. Hence, by the determination of the supersolubility curve, the calculated standard deviation is 2.24, which corresponds to temperature differences up to 5 °C. This could be due to the very low solubility of the samples, leading to increased discrepancies by the local formation of nuclei.

- Summary

In chapter 4.2, results from the solubilities and MSZW of aqueous solutions of the substances used in this thesis were represented and the dependency of the concentration on the temperature discussed. Generally, the solubility increases with the increase of the temperature. The solubility values for the single glutamic acid enantiomers are very close to those for aminobenzoic stereomers at lower temperatures (e.g. 0.3 - 0.4 wt.% at 10°C) and with the increase of the temperature the difference between the values rapidly increase up to 0.7 – 1.0 wt.% at 50°C. In comparison to them, the solubility of the asparagine enantiomers is almost 4 times higher with values from 1.4 wt.% at 10°C up to 7.6 wt.% at 50°C. This could be reasoned with the respected molecule structures (see Figs.

3.1, 3.2 and 3.3), where the benzyl core and the length of the carbon chain are the main reasons for the lower solubility of the aminobenzoic acid and glutamic acid in water respectively. On the contrary, shorter carbon chain of the asparagine along with hydrophilic functional groups (-COOH and -NH2) increases the molecule affinity to water and therefore the solubility of the substance. For glutamic acid and asparagine monohydrate systems incipient decomposition was suspected at temperatures above 65°C indicated by slight coloring of the solution samples, which resulted in a slight decrement of the measured values for the solid-liquid equilibria. The same trend was also observed in aminobenzoic acid system, starting at about 45°C. Based on the solubility measurements and the respected MSZW, it can be generally concluded that further crystallization experiments could be planned in the temperature region 30°C – 45°C with

Results and discussion Ch. 4

supersaturation up to 1.5, where the productivity could be relatively high, while attaining low nucleation probabilities and keeping the solutions stable with time.